Research Fellow at the Cambridge Crystallographic Data Centre. His research interests involve the use of structural information, mainly obtained by X-ray crystallography, to explain the activity of drugs and the properties of solid dosage forms. In periods of sabbatical leave at Brookhaven National Laboratory and Oxford University, he also applied neutron diffraction and computational chemistry. He has authored or co-authored three reviews on historical aspects of crystallography: the important pioneering crystallographic research done by Lars Vegard, renowned as a physicist studying the aurora borealis but little-known as a crystallographer; the identification by June Sutor of C-H…O hydrogen bonding, which met with disdain before eventually being confirmed; and the scientific and personal relationship between Max von Laue and the Braggs. Arthur Lindo Patterson, his function and element preferences in early crystal structures In 1934 Arthur Lindo Patterson showed that a map of interatomic vectors is obtainable from measured X-ray diffraction data without phase information. Such maps were interpretable for simple crystal structures, but proliferation and overlapping of peaks caused confusion as the number of atoms increased. Since the peak height of a vector between two particular atoms is related to the product of their atomic numbers, a complicated structure could effectively be reduced to a simple one by including just a few "heavy" atoms (of high atomic number) since their interatomic vectors would stand out from the general clutter. Once located, these atoms provide approximate phases for Fourier syntheses that reveal the locations of additional atoms. Surveys of small-molecule structures in the Cambridge Structural Database during the periods 1936-1969, when Patterson methods were commonly used, and 1980-2013, dominated by direct methods, demonstrate large differences in the abundance of certain elements. The "moderately heavy" elements K, Rb, As and Br are the heaviest element in the structure more than 3 times as often in the early period than in the recent period. Examples are given of three triumphs of the heavy atom method and two initial failures that had to be overcome.